Advent of Code is back! Unwrap daily challenges to sharpen your Alteryx skills and earn badges along the way! Learn more now.

Analytics

News, events, thought leadership and more.
Stephan
Alteryx
Alteryx

by Johnny Daly and Stephan Thodesen


An illustration of the need for analytical thinking

 

In the mid-'90s, Hans Rosling [1]  proved that medical students at the best university in Sweden knew statistically less than chimpanzees do about the state of global health. Interestingly enough, he also managed to prove that our primate cousins were roughly on par in this area with Professors at the Karolinska Institutet, which houses the Committee that gives out the Nobel Prize for medicine.

 

How can this be? How can some of the most intelligent people in the world get something that would seem relatively straightforward so wrong?

 

To rub salt in the wound, there was very little margin for interpretation of the right answer; of the 2 countries students and professors could choose between, one had twice the infant mortality rate of the other.

 

Which country has the highest child mortality in the mid-‘90s? [2] 

 

Sri Lanka or Turkey

Poland or South Korea

Malaysia or Russia

Pakistan or Vietnam

Thailand or South Africa

 

The reason that the illustrious Scandinavian academics, as well as the smartest students in one of the richest countries in the world, were unable to obtain better results than chimps lies somewhere on the continuum between cognitive bias and some form of psychological heuristics. Or, in plain English, we either have thought patterns that produce wrong answers, or our brains make best guesses based on earlier experiences that are not necessarily right. We might have potentially split some hairs here. These flawed reasoning styles were also the cause of particularly poor University exam performances for one of the authors as well. Extracurricular activities played absolutely no part whatsoever in that second scenario.

 

Depending on how widely or narrowly you interpret things (and this could be a very interesting, albeit potentially quite a niche philosophical debate), analytics are predominantly, if not exclusively, done to generate insights that will lead to decisions or the decision to not make a decision [3], which is, intriguingly enough, still decisive.

 

The quality of the analytics will affect the quality of the insights, which in turn will make the decision-making process more or less complicated. If not even the people who give out the prize that is commonly considered to be the global apex in terms of recognition in the field of medicine are immune, what chance do the rest of us stand against our own mind’s efforts to come up with the most time efficient (or lazy) answer that is not always right?

 

We actually stand a great one and have never been better placed in the history of humanity. This might sound like a rather bold and brazen statement. It might potentially be, but we don’t think so.

 

We believe the ultimate goal for people is to make a net positive contribution to the world. To do that, we need tools to look at how things are done today, understand them, and think of better ways they can be done tomorrow (see, we didn’t use decision again!). While there are multiple ways of doing the latter part (Kaizen, 6 Sigma, and the sadly less frequent “it came to me in a dream”), they will invariably start with a good understanding of what is actually going on.

 

Who is in charge of analytics?

 

Traditionally, the group of people who would handle this are analysts. A small group of well-trained people who understand how to ask the right questions to uncover the truth (or at least parts of it). If you are well organized and a little lucky, you are able to have some who also understand the broader area, so they can answer these questions faster. The challenge arises when you have lots of questions. While all of them are probably not equally important, they do need answers to be able to continue to move forward. What often ends up happening is the leaders will get the answers they need, but the rest of the group that wants to really work with insights will either not get them or only get the answers to questions the leadership interpret as important. While that will fulfill Pareto’s principle, there is a whole tail there that is left unlooked at.

 

The good news is that the traditional barriers to entry for white-collar jobs are more or less being dynamited every day. This is a good thing. It makes the pie bigger, and honestly, who doesn’t want more pie? And while analytics white collars might have the occasional caffeinated beverage stain or pizza fleck on them (although, in reality, how many analysts actually wear a white collar is a separate topic), we need more of them.

 

The sheer volume of information available, coupled with the hard work of user researchers across the world to make it easy to consume, and fit-for-purpose business models that make it accessible, ensure that pretty much anyone with a smartphone can learn pretty much anything, very often for free. Alteryx, for example, allows anyone to come to their community to learn how to use the software for free.

 

We need teams in factories to learn those same skills to make them more efficient and produce less greenhouse gases. We need teachers to learn those skills to improve the way they teach, making the next generation better than the one before, and as part of that, we need children to learn them so the way they look at challenges will be more nuanced than ours. It applies to pretty much any profession.

 

Why analytics should be democratized

 

Analytics, in essence, needs to become collarless.

 

While Simon Sinek has unfortunately nabbed the title for his book, we need to start with why and we need everyone to do it.  We need people to have the ability to ask questions and understand the answers. We need to ask why things are done the way they are done and if there is a better way to do it in the future. This is not new. Socrates went around doing it in ancient Greece some two and a half thousand years ago.

 

Unless we all have the ability to ask questions, we will always have the same answers, which sometimes are true, and sometimes are not. We will potentially be carrying forward the same preconceived notions and biases’, which will leave the world as it is. It is not a coincidence that the very word used for extracting information from a database is “to query.”

 

While the truth is often perceived differently from different vantage points, it is also indispensable to have a common language to discuss them. While Gloria Steinem is probably right when she says that the “Truth will set you free, but first it will piss you off,” we think that many people working to find the truth can sometimes just make all of us a bit happier without any temporary side effects.

 

Learn more about analytics democratization:

Alter Everything Episode 131: How democratization can boost your analytics maturity

How to Succeed with Analytics Democratization

 


[1] Should you ever feel the need for a little data driven optimism about the world around you, please read his book Factfulness: Ten reasons why we are wrong about the world – and why things are better than you think

[2] The TED talk The best stats you've ever seen is from 2006 and he said he did this experiment roughly 10 years earlier, so we went with mid 90’s

[3] “Decision” is the combination of 2 Latin words: “de” which means off and “caedere” which means to cut. So essentially it means to cut yourself off from options… Maybe not always the best thing to do?

Comments
georgermccrea
7 - Meteor

This is a fantastic and insightful blog.

 

I agree whole heartedly about the need to democratise analytics by creating the most diverse group of practitioners possible.  

The best analytics teams I’ve worked with have had a rich diversity and deliver great insight.